ggambetta

You can do rotations and translations "via linear algebra" without homogeneous coordinates (the 4-element tuple representing a 3d point or vector), since adding two vectors is linear algebra.

What this lets you do is composing multiple transforms into one matrix multiplication instead of a sequence of multiplications and additions; that's what dramatically increases performance, on modern computing devices but most especially on ancient ones, where we were fighting for every MUL.

More details: https://gabrielgambetta.com/computer-graphics-from-scratch/1...

I hope a similar document surfaces about Monkey Island (or even MI2).

Not a pirate but a Spanish speaker.

Interestingly,

He's a doctor [in general] -> el es médico

He's the ship's doctor [this specific ship] -> el es el médico del barco

He's a ship doctor [generally works in ships] -> el es médico de barco

Can't wait for the movie. I grew up on Roadrunner cartoons (among others, most notably Top Cat and Scooby Doo), and some of my earliest childhood drawings are elaborate diagrams of traps to catch the Roadrunner, with detailed if implausible Rube Goldberg mechanics. The Coyote might have nudged me towards engineering, not even joking.

Hence DiRT or whatever they call it these days

My book tries to be as accessible as possible; a high schooler with little to no knowledge of linear algebra should be able to understand all of it. So in every case I've chosen simplicity (code and/or conceptually) over performance, but without sacrificing correctness. Not always an easy balancing act!

I would not recommend anyone to write a production rasterizer using the algorithms in my book. This website seems more performance-focused (Bresenham is the obvious choice to draw lines, for example).

Specifically about clipping, you can ignore all the planes except the near plane, and everything will be fine. You'll try to draw more pixels outside the canvas but that's just slow, not problematic. The near plane does protect you from divisions by zero and from points with negative Z, which project upside down and mess things up.

Thanks for saying that :) Have you built something based on that idea? I still haven't, but I'm thinking on throwing some Claude at it for the most mechanical parts.

I have a whole chapter on that! https://gabrielgambetta.com/computer-graphics-from-scratch/1...

Yes, and that actually sharpens my previous conclusions.

That's the unlock!

So it's not a one-word theory, it's a theory with a one-word name... shocking. What a clickbait title!

They used to be called "Sales Engineers". They still are, but they used to, too.

I'm pretty sure I submitted this idea to the internal "file a patent" drive they were doing like 15 years ago. It was rejected. I thought it was either stupid or too obvious and they were already doing it...

For me it was "There’s a clean exception to all of this, and it flips the entire logic." and "here’s the reframe". That's Claude-speak.

No Striker? [0] Zero realism but incredibly fun to play. IIRC it had a two-player mode (sharing the keyboard) that was super fun.

[0] https://www.mobygames.com/game/12380/striker/

I was thinking exactly the same thing, this is CSP/SR. One of my favorite topics! https://gabrielgambetta.com/client-server-game-architecture....

I've read that. I'm not part of the conservator community, I just really enjoy the videos.

I fully agree with you. However, when non-LatAm people talk about "socialism", they probably think of European-style socialism, which is awesome. LatAm socialism is a lot more extreme, and closer to full-on communism - my country still has people stuck in the 60s cold war era mentality and fully supporting the ideals of of the Soviet Union, still praising Cuba and Venezuela, reality notwithstanding.

European-style socialism would be better. Full 60s socialism would be significantly worse. It's quite likely that people ITT are using "socialism" to refer to either, and we're all talking past each other.

They should just trap Julian Baumgartner inside for a month, with nothing but water, food, and conservation-grade varnish and reversible pigments, and the whole place would look like new!

It seems you're not from LatAm.

That's a great idea! Just haven't gotten that far yet (that article is pre-LLM, and I'm very early in a new AI-driven reconstruction). But yeah, taking the original extracted sprites and having Nano Banana upscale or redraw them is a great idea.

I'm very happy to see this! Not so much because of TDIII (which I played, although not nearly as much as Stunts), but because there seems to be some momentum building around recreating old games using AI agents, and I love that! I had explored some related ideas [0] but throwing Claude at the problem seems super promising. The recent Crimsonland thing [1] was great!

[0] https://www.gabrielgambetta.com/remakes.html

[1] https://banteg.xyz/posts/crimsonland/

Fair, I wouldn't have been able to write Bresenham back then (or now, off the top of my head). I'd have written a simple trig-based one. Maybe I'd have failed the interview :D

So at the end of each run increment the relevant pointer by (pitch - w) not pitch which I'm sure it's one of the bugs they saw all the time in this interview :)

Thought the same. Reckless today, literally standard back then.

This brings back memories! I could easily pass this interview today, because I used to write code like this all the time 25 years ago doing gamedev (and so did everyone else to some extent). But the really interesting thing is that I just realized I haven't written code like this in a long, long time.

Programming has changed over time, but the change has been so gradual I hadn't even realized this until this article. These days I'm pondering how the profession has changed in the last 2 years due to AI. Feels a lot more like a step change. And yet I'm having more fun than I've had in a long time, both at work and at home, throwing Claude at problems. I still don't fully understand why.

What I'm hearing is that the smoke in smoke tests comes from the oil tanks!

> And the public funded the research that made it possible. The transformer architecture,

Errr pretty sure that was Google?

The whole thing compiles with 2 warnings. Incredible codebase. John Carmack definitely was/is on a different level.

Back when I was making videogames I followed a similar philosophy. No warnings (but in an orders-of-magnitude smaller and less complex codebase). Crash on failed asserts, used liberally, in debug builds. Not sure why but it seems that gamedev doesn't do this kind of rigorous engineering in general (or at least it didn't back then -- and admittedly I never worked in a big studio).

Thanks for the detailed explanation, I had no idea about any of this.